Self-Tuning Transfer Dynamic Convolution Autoencoder for Quality Prediction of Multimode Processes With Shifts

Chao Yang, Qiang Liu*, Chen Wang, Jinliang Ding, Yiu ming Cheung Cheung

*Corresponding author for this work

Research output: Contribution to journalJournal articlepeer-review

Abstract

Process shift of multimode process involving data distribution and dynamic relation makes traditional transfer learning methods be intractable and even result in negative transfer. To tackle this issue, this article proposes a novel self-tuning transfer dynamic modeling method for quality prediction of multimode processes. First, in order to capture domain-invariant spatiotemporal (DIST) features, a transfer dynamic convolution autoencoder (TDCAE) with a feature decomposition structure is established. Meanwhile, a first-order vector autoregressive constraint is embedded to extract consistent inner dynamics for DIST features. Then, a shared regression network is established to extract the relations with quality variables. Furthermore, by making full use of private spatiotemporal information from target labeled samples in response to the process shift, the self-tuning TDCAE (STDCAE) aided by a fine-tuning strategy is established for online compensation. Finally, the efficacy of the proposed TDCAE and STDCAE is demonstrated by a comprehensive study of a three-phase flow facility process.

Original languageEnglish
Pages (from-to)11295-11305
Number of pages11
JournalIEEE Transactions on Industrial Informatics
Volume20
Issue number9
Early online date31 May 2024
DOIs
Publication statusPublished - Sept 2024

Scopus Subject Areas

  • Control and Systems Engineering
  • Information Systems
  • Computer Science Applications
  • Electrical and Electronic Engineering

User-Defined Keywords

  • Convolutional neural networks
  • deep autoencoder
  • dynamic process modeling
  • multimode processes
  • quality prediction
  • transfer learning (TL)

Fingerprint

Dive into the research topics of 'Self-Tuning Transfer Dynamic Convolution Autoencoder for Quality Prediction of Multimode Processes With Shifts'. Together they form a unique fingerprint.

Cite this